Instructions to use fastino/GLiDE with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use fastino/GLiDE with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="fastino/GLiDE")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("fastino/GLiDE")
model = AutoModelForMultimodalLM.from_pretrained("fastino/GLiDE", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use fastino/GLiDE with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "fastino/GLiDE"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "fastino/GLiDE",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/fastino/GLiDE

SGLang

How to use fastino/GLiDE with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "fastino/GLiDE" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "fastino/GLiDE",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "fastino/GLiDE" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "fastino/GLiDE",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use fastino/GLiDE with Docker Model Runner:
```
docker model run hf.co/fastino/GLiDE
```

GLiDE leads my 1,342-question System 1 benchmark, ahead of GPT-6 Sol

by etsabary - opened about 5 hours ago

Discussion

etsabary

about 5 hours ago

GLiDE is the first specialized decision model in my benchmark to outscore GPT-6 Sol with reasoning disabled, the highest-scoring LLM reference I have tested here.

All models answer the same 1,342 supplied-fact, five-choice questions across 31 categories. No tools or iterative problem-solving.

Model	Correct	Accuracy	Median response	Cost for all 1,342
Fastino GLiDE	1,154	86.0%	213 ms	~$0.0174*
GPT-6 Sol, reasoning off	1,143	85.2%	1,023 ms	$0.8483
Mammoth-1	1,139	84.9%	240 ms	Promotional free access

GLiDE's measured median was 4.8 times shorter than Sol's, at about 1/49 of the run cost. The accuracy lead is 11 answers, or 0.82 percentage points. This is a developing benchmark, not evidence of universal intelligence superiority; dates, providers and server loads differ.

Some category results are particularly interesting:

Truth-tellers and liars: 59/74, versus Sol's 38/74. These questions require finding which assignments of honest and dishonest speakers fit their statements.
Rule violations: 52/52. Required conditions: 51/52.
Ordered procedures: 18/48, versus Respan Span-02's 33/48. Applying successive edits to the current state remains difficult.

On the separate 200-question image benchmark, GLiDE scored 158/200 (79.0%), second to Amos at 159/200.

Tested October 23, 2026 through Fastino's hosted fastino/GLiDE alias; responses identified glide_v2. The live catalog described v2.2. This is not an exact-weight identification of the downloadable model. Sol's comparison run is from October 19.

*Fastino cost is calculated from reported input tokens at the captured $0.04/million-input-token tariff, not returned monetary receipts. Promotional free access is not a production tariff.

Full text report, category scores and methodology · Questions and reproducible evidence · Image report

Sahibzada-a

fastino org about 5 hours ago

Really love this work, Prof. @etsabary !

etsabary

about 4 hours ago

Thank you so much, Sahibzada! I’m really impressed by GLiDE. Its combination of accuracy, speed and cost is remarkable, and the truth-teller results were especially striking in my tests. Congratulations to you and the Fastino team, and thank you for making it available for independent testing!

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment